1 Reply Latest reply on Feb 16, 2011 1:37 PM by mrfixit440

    SpyJMSException in 4.2.3.GA using postgresSQL 1.8.4

    mrfixit440

      this stacktrace happens after a few hours everyday.  Not sure what to try. Also, don't understand why a non persistant topic is being written to the jms_messages table in the first place.  Also, why is it always this topic and not any others?  I see it says an I/O error to the DB, was wondering if there is some configuration on jboss side for retries? I see in the postgres log, there is 4 successful commits during the same timeframe (during the same second).  Any insight or ideas would be greatly appreciated.

       

      org.jboss.mq.SpyJMSException: Could not remove message: 325 msg=30 hard STORED NON_PERSISTENT queue=TOPIC.JemLoggerTopic.ID:4.-2147483648 priority=4 lateClone=false hashCode=21428928; - nested throwable: (org.postgresql.util.PSQLException: An I/O error occured while sending to the backend.)

                at org.jboss.mq.pm.jdbc2.PersistenceManager.removeFromStorage(PersistenceManager.java:1617)

                at org.jboss.mq.server.MessageCache.removeFromStorage(MessageCache.java:425)

                at org.jboss.mq.server.MessageReference.clear(MessageReference.java:289)

                at org.jboss.mq.server.MessageCache.removeInternal(MessageCache.java:212)

                at org.jboss.mq.server.MessageCache.remove(MessageCache.java:169)

                at org.jboss.mq.server.BasicQueue$RemoveMessageTask.run(BasicQueue.java:1444)

                at org.jboss.mq.pm.TxManager.addPostCommitTask(TxManager.java:174)

                at org.jboss.mq.server.BasicQueue.performOrPrepareAcknowledgeMessage(BasicQueue.java:806)

                at org.jboss.mq.server.BasicQueue.acknowledge(BasicQueue.java:777)

                at org.jboss.mq.server.JMSTopic.acknowledge(JMSTopic.java:362)

                at org.jboss.mq.server.ClientConsumer.acknowledge(ClientConsumer.java:348)

                at org.jboss.mq.server.JMSDestinationManager.acknowledge(JMSDestinationManager.java:538)

                at org.jboss.mq.server.JMSDestinationManager.acknowledge(JMSDestinationManager.java:531)

                at org.jboss.mq.server.JMSServerInterceptorSupport.acknowledge(JMSServerInterceptorSupport.java:131)

                at org.jboss.mq.server.TracingInterceptor.acknowledge(TracingInterceptor.java:385)

                at org.jboss.mq.server.JMSServerInvoker.acknowledge(JMSServerInvoker.java:137)

                at org.jboss.mq.il.uil2.ServerSocketManagerHandler.handleMsg(ServerSocketManagerHandler.java:108)

                at org.jboss.mq.il.uil2.SocketManager$ReadTask.handleMsg(SocketManager.java:419)

                at org.jboss.mq.il.uil2.msgs.BaseMsg.run(BaseMsg.java:398)

                at EDU.oswego.cs.dl.util.concurrent.PooledExecutor$Worker.run(PooledExecutor.java:761)

                at java.lang.Thread.run(Unknown Source)

      Caused by: org.postgresql.util.PSQLException: An I/O error occured while sending to the backend.

        • 1. SpyJMSException in 4.2.3.GA using postgresSQL 1.8.4
          mrfixit440

          Anyone have any ideas?  To clarify issue, we've setup up postgreSQL to be our persistant storage of JMS queues.  One non-persistant topic is getting this error.  I see the MessageCache's HighMemoryMark is set to 50 MB's.  I guess my JVM memory usage is getting close to this value, and thats why my non-persistant topics are getting written to the DB in the first place.  This is hard to believe, has anyone seen a similiar error?